Skip to content

atumanov/ray-scheduler-prototype

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Ray Scheduler Prototype

Overview

Aims of the scheduler prototype include the following:

  • Same APIs as a production scheduler
  • Easy to add new algorithms
  • Able to replay production logs
  • Support unit tests for core scheduler and for database (Redis) integration

Usage Example

(NUM_NODES=3
NUM_WORKERS_PER_NODE=1
for trace in $(ls traces/test/*.json); do
    python replaytrace.py $NUM_NODES $NUM_WORKERS_PER_NODE 0.001 .0001 trivial fifo true $trace;
done)

Installation

  • install anaconda
  • pip install ortools numpy

About

Experimental code for Ray scheduler evaluation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • OpenEdge ABL 98.6%
  • Python 1.4%